Energy consumption and water production cost of conventional and renewable-energy-powered desalination processes

Desalination technologies improve water quality, greatly reduce water shortage problems, and improve quality of life and economic status. Two main technologies are currently used in water desalination: thermal (phase-change) processes and membrane processes. The primary thermal distillation processes include multistage flash distillation (MSF), multi-effect distillation (MED), and vapor compression (VC). The VC process encompasses two types: mechanical (MVC) and thermal (TVC). The common membrane desalination processes include reverse osmosis (RO) and electrodialysis (ED and EDR). Energy cost, operational and maintenance cost, and capital investment are the main contributors to the water production cost of any of these processes. The energy cost is responsible for about 50% of the produced water cost. For thermal distillation processes (MSF, MED, and TVC), two energy forms are required for the operation: (1) low-temperature heat, which represents the main portion of the energy input and is usually supplied to the system by a number of external sources (e.g., fossil fuel, waste energy, nuclear, solar) and (2) electricity, which is used to drive the system’s pumps and other electrical components. For the MVC thermal distillation process, only electricity is needed. For membrane processes (RO and ED), only electricity is required as an energy input. Renewable energy systems such as solar thermal, solar photovoltaic, wind, and geothermal technologies are currently used as energy suppliers for desalination systems. These renewable resources are now a proven technology and remain economically promising for remote regions, where connection to the public electric grid is either not cost effective or feasible, and where water scarcity is severe. As the technologies continue to improve, and as fresh water becomes scarce and fossil fuel energy prices rise, renewable energy desalination becomes more viable economically. The technical features, energy consumption, environmental considerations, and potential of renewable energy use in driving the main desalination processes are reviewed and analyzed in this paper. The current and projected costs of water produced from conventional and renewable-energydriven processes are discussed and compared. & 2013 Elsevier Ltd. All rights reserved. 1. Conventional desalination processes Desalination technologies are categorized as thermal (phasechange) and membrane desalination, and these are further divided into subgroups. The main thermal distillation technologies are multi-stage flash (MSF), multi-effect distillation (MED), vapor compression (VC), whereas the main membrane technologies are reverse osmosis (RO) and electro-dialysis (ED and EDR) [1]. 1.1. Thermal distillation technologies 1.1.1. Multi-stage flash MSF distillation is an energy-intensive process that requires both thermal and electrical energy. The thermal energy is in the form of low-pressure bleed steam (1 to 3 bars) for the feed-brine heating, and medium-pressure steam for the ejectors to generate the required vacuum in different sections of the unit. The electrical energy is required for driving the unit’s various pumps such as recycle, cooling water, distillate product, brine blow down, condensate, and chemical dosing pumps. MSF units typically range from 10,000 to 35,000 m/day and consist of a series of stages, ranging from 4 to 40 each, with successively lower temperature and pressure that cause flash evaporation of the hot ll rights reserved. brine followed by condensation as fresh water. In this process, the feed seawater moves in heat exchangers through the stages and gains some heat that helps to reduce the external thermal energy needed for hot brine and also to condense the water vapor for collection as fresh water in each stage. External heat from fossilfuel boilers, power-plant waste heat, nuclear reactor, renewable energy, or any other heating source is supplied to the intake preheated seawater to raise its temperature to the required top brine temperature of 901 to 110 1C. The heated brine water is then moved through stages that are held at successively lower pressure in which a small amount of water flashes to vapor in each stage and the remaining brine flows to the next stage for further flashing until it is finally discharged. The vapor from each stage is condensed and collected as fresh water [2,3]. Fig. 1 shows a schematic diagram of the MSF unit. Flashing of the steam forms scales and deposits on the tubes, so periodic cleaning and removal is required. MSF is currently the second-most desalination process installed worldwide after the RO process. 1.1.2. Multi-effect distillation The MED process consists of a series of stages (usually from 2 to 16) that are maintained at decreasing levels of pressure. Fig. 1. Schematic diagram of MSF unit. Fig. 2. Schematic diagram of MED unit. Letter to the Editor / Renewable and Sustainable Energy Reviews 24 (2013) 343–356 344 External heat from a fossil-fuel boiler, power-plant waste heat, solar, or other sources is supplied to increase the brine temperature of the first stage to around 70 1C, to be used to evaporate some of the brine inside the stage that is kept at low pressure. The water vapor produced from the stage is transferred inside a tube to the next heating stage for boiling additional seawater, which produces water vapor in a series fashion. MED units are generally built at capacities of 600 to 30,000m/day and the design is based on two arrangements: the vertical tube in which the seawater boils in a thin film flowing inside the tube and vapor condenses on the heat-transfer tubes, and horizontal tube where the seawater feed is sprayed on the outer surface of the tubes and vapor flows inside the horizontal tubes, where it condenses to produce water. Fig. 2 shows a schematic diagram of the MED unit. The earliest distillation plants used MED, but MSF displaced it due to its lower cost and less tendency to scale [4]. In the past few years, the interest in the MED process has been renewed and appears to be gaining market share. 1.1.3. Mechanical vapor compression Distillation plants using vapor compression rely on the heat generated by the compression of water vapor to evaporate salt water, and two methods are employed—mechanical vapor compression (MVC) and thermo vapor compression (TVC). The feed water enters the VC process through a heat exchanger, and vapor is generated in the evaporator and compresses by mechanical (MVC) or thermal (TVC) means. Compression the vapor raises its temperature by a sufficient amount to serve as the heat source. The concentrated brine is removed from the evaporator vessel by the concentrate reticulating pump. This flow is then split, and a portion is mixed with the incoming feed and the remainder is pumped to the waste. Fig. 3 show both types. MVC use electricity to drive the compressor, whereas in TVC a steam jet creates the lower pressure. These units are usually used in smalland medium-sized applications. MVC capacity ranges between 100 and 3000m/day, and TVC capacity ranges between 10,000 and 30,000m/day [5]. 1.2. Membrane desalination technologies 1.2.1. Reverse osmosis Reverse osmosis (RO) is a form of pressurized filtration in which the filter is a semi-permeable membrane that allows water, but not salt, to pass through. This yields permeated fresh water and leaves a concentrated solution on the high-pressure side of the membrane. It has four subsystems: (1) pre-treatment, (2) high-pressure pump, (3) membrane, and (4) post-treatment. Feed-water pre-treatment involves filtration, sterilization, and addition of chemicals to prevent scaling and biofouling. The high-pressure pump generates the pressure needed to force the water to pass through the membrane; therefore, the energy needed is electricity to drive the pumps. The pressure needed for desalination ranges from 17 to 27 bars for brackish water and from 55 to 82 bars for seawater. The membranes are designed to yield a permeate water of about 500 ppm and made in a variety of configurations. Several types of membrane are available Fig. 3. Schematic diagram of VC (MVC and TVC) units. Fig. 4. Schematic diagram of RO system. Letter to the Editor / Renewable and Sustainable Energy Reviews 24 (2013) 343–356 345 in the market, with the two most commonly used ones being spiralwound and hollow fine fiber. The post-treatment removes gases such as hydrogen sulfide and adjusts pH. Fig. 4 is a schematic diagram of an RO system. RO is a mature technology that is the most commonly used desalination technique. Its installed capacity ranges between 0.1 m/day (used in marine and household applications) to 395,000m/day (for commercial applications) [6–12]. Fig. 5. Schematic diagram of ED unit. 1.2.2. Electro-dialysis and electro-dialysis reversal Electrodialysis (ED) is an electrochemical separation process that operates at atmospheric pressure and uses direct electrical current to move salt ions selectively through a membrane, leaving fresh water behind. The ED unit consists of the following components: pretreatment system, membrane stack, low-pressure circulation pump, direct-current power supply (rectifier or photovoltaic system), and post-treatment system. The operational principle of ED is as follows: electrodes (generally constructed from niobium or titanium with a platinum coating) are connected to an outside source of direct current in a container of salt water containing an ionselective membrane connected in parallel to form channels. When brackish water flows between these channels and electricity is charging the electrodes, positive salt ions travel through the cationpermeable membrane toward negative electrodes, and negative salt ions travel through the anion-permeable membrane to the positive electrode, which results in the removal of salinity from the water. This creates alternating channels—a concentrated channel for the brine and a diluted channel for the product fresh water [9,12]. An ED plant’s typical capacity ranges from 2 to 145,000m/day. Fig. 5 shows the schematic diagram of an ED unit. In EDR, the polarity of the electrodes is switched periodically. The concentrate stream is then converted to the feed stream and the feed stream becomes the concentrate stream. Reversing the flow increases the life of the electrodes and helps to clean the membranes. When the membranes Letter to the Editor / Renewable and Sustainable Energy Reviews 24 (2013) 343–356 346 are operated in the same direction all the time, precipitant can build up on the concentrate sides [13]. 2. Energy requirement for desalination processes 2.1. Minimum energy requirement for desalination All desalination processes are energy intensive and share a common minimum energy requirement for driving the separation of a saline solution into pure water and concentrated brine. It is independent of the detailed technology employed, exact mechanism, or number of process stages. The concept of minimal energy for the separation process is well established in thermodynamics. The solute movement is wholly determined by fluctuations of thermal collisions with nearby solvent molecules. The minimum work needed is equal to the difference in free energy between the incoming feed (i.e., seawater) and outgoing streams (i.e., product water and discharge brine). Different methods were used to calculate the minimum energy requirement of water desalination. Using the van’t Hoff formula for normal seawater of salinity equal to 33,000 ppm at 25 1C, the minimumwork has been calculated as 0.77 kW h/m [14]. 2.2. Actual energy requirement of main desalination processes The actual work required is likely to be many times the theoretically possible minimum. This is due to the extra work required to keep the process going at a finite rate, rather than to achieve the separation. Currently, desalination plants use 5 to 26 times as much work as the theoretical minimum, depending on the type of process used. Due to this intensive energy consumption, there is a need to make desalination processes as energy efficient as possible by improving the technology and economies of scale. RO, ED, and VC systems use electricity as a primary source of energy, whereas MSF, MED, and TVC systems use thermal energy as a primary source and electricity to drive associated pumps as a secondary source. Electricity could be generated from fossil fuel (coal, oil, and gas), renewable energy, and nuclear sources. Thermal energy could be produced from fossil-fuel-fired boilers, power-plant waste heat, renewable energy sources, and industrial-waste heat sources. 3. Energy consumption of the main processes 3.1. Distillation processes Two types of energy – low-temperature heat and electricity – are required for most distillation processes (MSF, MED, and TVC). The low-temperature heat represents the main portion of the energy input and the electricity is used to drive the system’s pumps. For the MVC process, only electricity is required. All thermal processes are equipped with condenser-tube bundles and numbers of large pumping units, including pumps for seawater intake, distillate product, brine blow down, and chemical dosing. The simplest distillation technique, single-stage evaporation, consumes a tremendous amount of energy. Boiling water requires around 650 kW h/m of product, depending slightly on the evaporation temperature. The main evaporation techniques (MSF and MED) have overcome this obstacle by reusing the energy consumption through multiple stages. The efficiency of the low-temperature heat is usually identified by one of two equivalent parameters: (1) the gain output ratio (GOR), which is a measure of how much thermal energy is consumed in the desalination process, and is defined as the ratio of the mass of distillate (kg) to the mass (kg) of the input steam, and (2) the performance ratio (PR), which is the mass of distillate (kg) per 2326 kJ. 3.1.1. Energy consumption in MSF process The energy consumption of the MSF depends on several factors: maximum temperature of the heat source, temperature of the heat sink, number of stages, salt concentration in the flashing brine solution, geometrical configuration of the flashing stage, construction materials, and design configuration of heatexchange devices. Therefore, the energy consumption of the MSF unit can be reduced by increasing the GOR (or PR), number of stages, and the heat-transfer area [15–21]. The MSF process operates at a top brine temperature (TBT) in the range of 901 to 110 1C. An increase of TBT increases the flash range, which, in turn, increases the production rate and improves the performance. However, selection of TBT is limited by the temperature to which the brine can be heated before serious scaling occurs. MSF commercial manufacturers provide a GOR design range between 8 and 12 kgdistillate/kgsteam depending on the steam feed temperature [15]; but the reported typical GOR in the Arab Gulf countries’ plants ranges between 8 and 10, and the typical PR ranges between 3.5 and 4.5 kgdistillate/MJ [16]. If we use the manufacturers’ values, then the thermal energy consumption of an MSF plant ranges between 190 MJ/m (GOR1⁄412) and 282 MJ/ m (GOR1⁄48). The electrical energy equivalent to these values based on power plant efficiency of 30% ranges between 15.83 and 23.5 kW he/m. The electricity consumption of the pumps ranges between 2.5 and 5 kW he/m therefore, the total equivalent energy consumption of the MSF unit ranges between 19.58 and 27.25 kW he/m. 3.1.2. Energy consumption in MED process The MED process also requires two types of energy—lowtemperature heat for evaporation and electricity for pumps. It operates at brine temperatures ranging from 641 to 70 1C. The manufacturers of MED units provide a GOR design ranging from 10 to 16. Typical Arab Gulf countries’ MED plants operate at lower GOR values of 8 to 12 [16]. If we use the manufacturers’ values, then the thermal energy consumption of MED plants ranges between 145 MJ/m (GOR1⁄416) to 230 MJ/m (GOR1⁄410). The work equivalent to these values based on a power-plant efficiency of 30% ranges from 12.2 to 19.1 kW he/m. The total electricity consumption of the pumps ranges from 2.0 to 2.5 [15]; therefore, the total equivalent energy consumption of the MSF units ranges from 14.45 to 21.35 kW he/m. 3.1.3. Energy consumption in MVC and TVC processes MVC needs electrical or mechanical energy only. It operates at a maximum TBT around 74 1C, with electrical energy consumption ranging from 7 to 12 kW he/m [15]. For TVC, both lowtemperature heat and electricity are needed. At TBT ranges from 631 to 70 1C, GOR of around 12, a heat input of 227.3 MJ/m (14.56 kW he/m), and electricity consumption of 1.6–1.8 kW he/ m are required [15]. Therefore, the total energy consumption of the TVC process is about 16.26 kW he/m. 3.2. Membrane processes Electricity is the only form of energy consumed in the membrane processes. For the RO process, AC electricity is consumed to drive the different pumps, whereas DC electricity is consumed in the ED electrodes and AC or DC electricity is consumed to drive the ED pumps. Letter to the Editor / Renewable and Sustainable Energy Reviews 24 (2013) 343–356 347 3.2.1. Energy consumption in RO process Electricity is the only required form of energy in the RO process. Energy consumption of the RO unit depends mainly on the salinity of the feed water and the recovery rate. The osmotic pressure is related to the total dissolved solids (TDS) concentration of the feed water; therefore, high-salinity water requires a higher amount of energy due to higher osmotic pressure. RO unit sizes vary from a very small unit with a capacity of 0.1 m/day to a 395,000 m/day plant. The average reported energy consumption ranges from 3.7 to 8 kW h/m [15,21,22]. The consumption may exceed 15 kW h/m for very small sizes units. For a typical size of seawater RO (SWRO) unit of 24,000 m/day, the electricity consumption ranges from 4 to 6 kW h/m with an energy recovery (ER) system for seawater. Low pressure is needed to desalinate brackish water; therefore, different membranes are used and much higher recovery ratios are possible, which makes energy consumption low. For a brackish-water RO (BWRO) unit, the electrical energy consumption ranges from 1.5 to 2.5 kW h/m